Optimal forest harvest age considering carbonsequestration in multiple carbon pools: a comparative

statics analysis

Patrick Asante1, Glen W. Armstrong1,

aDepartment of Renewable ResourcesUniversity of Alberta

Edmonton, Canada T6G 2H1bDepartment of Renewable Resources

University of AlbertaEdmonton, Canada T6G 2H1

Abstract

We present an analytical model for determination of the economically optimal

harvest age of a forest stand considering timber value, and the value of carbon

fluxes in living biomass, dead organic matter, and wood products pools. Through

comparative statics analysis, we find that consideration of timber value and fluxes

in biomass carbon increase harvest age relative to the timber only solution, and

that the effect on optimal harvest age of incorporating fluxes in the dead organic

matter and wood products pools is indeterminate.

We also present a numerical example to examine the magnitudes of these ef-

fects. In general, incorporating the dead organic matter and wood products pools

have the effect of reducing rotation age. Perhaps more interestingly, when initial

stocks of carbon in dead organic matter or wood products pool is relatively high,

consideration of these pools can have a highly negative effect on net present value.

∗Corresponding author.

Preprint submitted to Journal of Forest Economics March 14, 2012

Keywords: optimal rotation, boreal forest, carbon market

2

1. Introduction

Concerns about climate change associated with increasing concentrations of

greenhouse gases (GHGs) has led to addition of carbon sequestration to the list

of ecosystem services provided by forests with potential economic value. Forests

may mitigate the effects of greenhouse gas (GHG) induced change, since trees

remove substantial amounts of CO2 from the atmosphere through photosynthesis

(IPCC, 2000). It is important to remember, however, that they also release CO2 to

the atmosphere through the processes of respiration and decomposition.

The choice of harvest age is the fundamental decision in an even-aged silvi-

cultural system. Generally, delaying the harvest age allows stands to increase in

volume, thereby storing more carbon (Harmon and Marks, 2002). Forests man-

aged on a longer harvest cycle accumulate more dead organic matter (DOM) and,

on average, store more carbon than forests managed on a shorter harvest cycle

(Krankina and Harmon, 2006). The choice of harvest age will also affect the stock

of carbon stored in wood products. If wood products decompose at a slower rate

than DOM in the forest, there may be an advantage to choosing a harvest age that

provides a larger mean annual increment (MAI), and therefore more wood prod-

ucts. There may also be some benefits of substituting wood products for more

GHG intensive construction materials such as concrete and steel.

Hoen (1994), van Kooten et al. (1995), Hoen and Solberg (1997), and Stain-

back and Alavalapati (2002) have investigated the impact of carbon tax and sub-

sidy schemes on the optimal harvest age for even-aged management. In these

models landowners are paid a subsidy for periodic carbon uptake in living biomass

and taxed when carbon is released through harvest or decay. The models devel-

oped in the above studies are essentially variations on the Hartman (1976) model,

3

which includes non-timber benefits related to the standing forest. These models

demonstrate that carbon taxes and subsidies will affect the optimal forest harvest

age and, consequently, the carbon stored in forests. Englin and Callaway (1995)

examined the impact of carbon price on the optimal harvest age of Douglas-fir and

concluded that rotation age increases with increasing carbon price. Plantinga and

Birdsey (1994) used an analytical model in the framework of Hartman (1976) to

show that with carbon benefits only in the analysis, the harvest age was infinite

in most cases and with carbon and timber, the harvest age would be between the

carbon only and the timber only harvest age. Enzinger and Jeffs (2000) showed

that the optimal rotation for Eucalyptus spp. plantations is longer when carbon

payments are considered than when they are not.

These studies do not account for carbon stored in dead wood, the forest floor,

soil, or wood products. A substantial proportion of the total carbon stored in forest

stands is dead organic matter. The choice of harvest age can have a substantial

effect on soil carbon stocks (Aber et al., 1978; Kaipainen et al., 2004). In Asante

et al. (2011), we demonstrated that the optimal harvest age can differ substantially

between cases where carbon in the DOM pool is considered and where it is not.

It may also be important to consider carbon stored in the wood product pool. The

amount of carbon in the wood product pool is affected by the choice of harvest

age as it directly affects harvest volume. The harvest age also affects the size

and distribution of logs in the stand and therefore the product mix (Krankina and

Harmon, 2006).

The study presented here differs from the previous studies because it presents

a matheatical model of the economics of forest carbon sequestration suitable for

a comparative statics analysis, considering timber values and carbon values for

4

biomass, dead organic matter, and wood product pools. The effect of changing

carbon prices and the inclusion of different pools in the analysis is considered. A

numerical example is also presented.

2. The Analytical Model

The analytical model assumes that the landowner wishes to determine the op-

timal harvest age, T ∗, that maximizes the net present value (NPV) of timber and

carbon sequestration values of an area of bare forest land. To simplify this analy-

sis, it is also assumed that the forest is managed under a single cycle establishment,

growth, and harvest. We make this assumption because as we showed numerically

in (Asante et al., 2011), the optimal harvest age may be dependent on the carbon

stocks in the DOM pool and this will change over time: it is possible that the op-

timal harvest age for any given rotation is different from the optimal harvest age

for subsequent rotations (see Fig. 1). A correctly formulated infinite time horizon

model should incorporate the possibility for multiple rotation ages: unfortunately,

our mathematical skills are not up to that challenge. Nonetheless, we believe

something interesting can be learned from a single rotation model.

A reviewer suggested that we could use a generalized Faustmann model such

as that used by Chang (1998) and by Armstrong and Phillips (1989) to account for

the value of future stands. This technique is used to find the optimal harvest age for

the first rotation, given an exogenous specification of the land expectation value

(LEV) of the stand at the beginning of the second rotation. The generalized Faust-

mann model is a useful way of thinking about the optimal forest rotation problem,

given the near certainty that “market conditions, government policies, regenera-

tion technology, management activities, and timber stand growth can change the

5

land expectation value from timber crop to timber crop” (Chang 1998, p.654).

This technique would work well if the future LEV can be viewed as exogenous.

However, in the model we present here, future LEV is related directly to the har-

vest age chosen for the current stand, as that decision affects the initial dead or-

ganic matter and product pool stocks for subsequent rotations, and consequently

affects the LEV. For the purposes of this paper, we believe a correct specifica-

tion of a single period model is more useful than an incorrect specification for an

infinite time horizon model.

2.1. Timber only

We will begin by building the timber only model. All prices and costs are

expressed in Canadian dollars (CAD). The NPV of one rotation of a stand of

timber is

NPVt = ((Pw −Cv) V[T ] −Ca) e−ρt −Ce (1)

where Pw is the price of timber (CAD/m3), Cv is the harvest cost per unit volume

(CAD/m3), Ca is the fixed harvest cost per unit area (CAD/ha), Ce is the stand

establishment cost (CAD/ha), and ρ is the real discount rate. The volume of timber

(m3 ha−1) standing on the forest at any age t is given by the timber yield function

V[t]. The timber harvest volume of a stand harvested at age T is given by V[T ].

The first order condition for maximization is

NPV′t[t] = ρ (((Pw −Cv) V[T ]) −Ca) − (Pw −Cv) V ′[T ] = 0 (2)

which can be rearranged to become: Find T such that

ρ =(Pw −Cv) V ′[T ]

(Pw −Cv) V[T ] −Ca (3)

6

The right hand side of Eq. 3 will be referred to as the relative value growth rate

(RVGR). The harvest age should be chosen such that the RVGR at the harvest

age is equal to the discount rate. We will find it useful to create an alias for the

numerator, Xt = (Pw −Cv) V ′[T ], and the denominator, Y t = (Pw −Cv) V[T ]−Ca.

Using the timber yield and financial parameters from (Asante et al., 2011), the

decision rule is illustrated in Fig. 2.

We will use the simple model in Fig. 2 as the basis for the comparative statics

analysis of the effect of including different carbon pools in the optimization. If

the RVGR curve shifts upward, the optimal harvest age is older. Conversely, if the

RVGR curve shifts downward, the optimal harvest age is younger.

2.2. Timber plus biomass pool

Suppose there exists a market where a forest landowner is paid for an increase

in the mass of carbon stored in her trees and pays for a decrease. The NPV of

biomass for carbon sequestration in one timber rotation can be written as

NPVb =

∫ T

0e−ρtPcB′[t] dt − e−ρT PcB[T ] (4)

where Pc is the price of carbon (CAD per tonne of carbon (tC henceforth)), and

B[t] expresses the mass of carbon in living trees (biomass, tC/ha) as a function of

stand age, t (years). The term to the left of the minus sign expresses the NPV of

carbon stored over the life of the stand; the term to the right expresses the payment

that must be made when the stand is harvested, as the living biomass is assumed

to be set to zero at the time of harvest.

The NPV of timber harvest and carbon sequestration services from biomass

can be calculated as

NPVtb = NPVt + NPVb (5)

7

The first order condition for maximization is

NPV′tb = e−ρt (−Caρ + PcρB[T ] + (Cv − Pw)(ρV[T ] − V ′[T ]

))= 0 (6)

which can be rearranged to become: Find T such that

ρ =(Pw −Cv) V ′[T ]

(Pw −Cv) V[T ] −Ca − PcB[T ]. (7)

If we substitute in the aliases for the numerator and denominator from the timber

only model, we get

ρ =Xt

Y t − PcB[T ]. (8)

B[T ] is assumed to be increasing with T . For positive carbon prices, the right hand

side of Eq. 8 is always larger than the right hand side of Eq. 3. This means that

the relative growth rate curve for timber and biomass is always above the relative

growth rate curve for timber only. For any discount rate, the optimal harvest age

will be older when timber and biomass are considered as compared to the timber

only case. One way of interpreting Eq. 8 is that the optimal harvest age is older in

order to delay the penalty associated with removing the biomass pool at harvest.

We will find it useful to create an alias for the denominator of Eq. 8, Y tb =

Y t − PcB[T ], so that we can compare solutions considering other carbon pools to

the timber plus biomass solution.

2.3. Timber plus biomass and dead organic matter (DOM) pools

As we did in Asante et al. (2011), we assume that the DOM pool in the model

decays at a constant rate, α. The living biomass contributes to the DOM pool

through a process we call litterfall. A fixed proportion, β, of the biomass pool is

8

added to the DOM pool. The change in the DOM pool is described as a differential

equation

D′[t] = βB[t] − αD[t]. (9)

If we solve this differential equation for D[t] given an initial DOM stock of D[0],

the DOM stocks at any stand age can be written as

D[t] = e−tα

(D[0] +

∫ t

0ekαβB[k] dk

). (10)

The rate of change in DOM stock, given D[0] is

D′[t] = βB[t] − e−tαα

(D[0] +

∫ t

0ekαβB[k] dk

). (11)

It is important to note that this rate of change is dependent on the initial stock,

D[0], the the DOM pool.

We assume that the market rewards accumulation of carbon in the DOM pool

and penalizes reductions. The NPV of the DOM pool is

NPVd =

∫ T

0PcD′[t]e−ρt dt + Pc(B[T ] − γV[T ])e−ρT (12)

The term to the left of the plus sign shows the contribution to NPV resulting from

the DOM decay and litterfall processes. The term to the right of the plus sign

indicates the value of the pulse of input to the DOM pool associated with harvest.

The parameter, γ is a conversion factor used to convert timber volume (m3) to

mass of carbon (tC). All biomass, except that removed as merchantable volume,

is transferred to the DOM pool at the time of harvest.

The NPV of timber plus the biomass and DOM pools is

NPVtbd = NPVt + NPVb + NPVd (13)

9

and the corresponding harvest age decision rule, with the aliases for the numerator

and denominator from the timber only solution substituted in, is

ρ =Xt + Pc (B′[T ] − γV ′[T ]) + PcD′[T ]

Y t − PcγV[T ]. (14)

The incorporation of −PCγV[T ] will have the effect of lengthening the rotation to

delay the loss of the harvested wood volume to the system. However, the direction

of change implied by the numerator is indeterminate. The term Pc (B′[T ] − γV ′[T ])

will generally be positive, as it represents the difference between the rate of growth

of biomass and the biomass equivalent of merchantable volume (one possible ex-

ception to this general statement could occur at a stand age when many of the

trees in reach a merchantability threshold: at this age the rate of increase in mer-

chantable biomass good be greater than the rate of increase of total biomass).

However, PcD′[T ], could be positive or negative depending on stand age, decay

rate, litterfall rate, and initial DOM stocks, D[0]. All other things being equal, a

higher D[0] will result in a lower rotation age.

When we substitute in timber plus biomass solution,

ρ =Xt + Pc (B′[T ] − γV ′[T ]) + PcD′[T ]

Y tb + Pc (B[T ] − γV[T ]), (15)

the discussion relating to the numerator remains the same, so the conclusions

about an indeterminate direction must hold. However as compared to the timber

plus biomass solution, the denominator will always be larger, which results in a

shorter rotation relative to the timber plus biomass solution.

The numerator for Eq. 15 will be aliased as Xtbd = Xt + Pc (B′[T ] − γV ′[T ]) +

PcD′[T ] and the denominator by Y tbd = Y tb + Pc (B[T ] − γV[T ]) for subsequent

discussions.

10

2.4. Timber plus biomass, DOM, and wood product pools

Carbon is stored in wood products, and the amount of wood products pro-

duced is related to the volume of trees harvested, and volume is assumed to be

increasing with stand age. Wood products are considered in the forest carbon

offsets protocol developed by California Climate Action Registry (2009). In the

California Climate Action Registry protocol, offset credits are increased with in-

creasing wood product production and decreased with decreasing wood product

production.

Including the wood product pool in the calculation of carbon offsets paid to

a forest landowner is difficult to justify, in our minds. It is unlikely that the

landowner has any custody of the wood once it leaves her forest, or perhaps

her mill. The choice of final end use for the products is out of the hands of the

landowner. However, there is at least one existing protocol that incorporates for-

est products, so for completeness, we will include forest products pool in this

analysis.

Because our analysis is at the stand level, we will define a forest products

pool that will represent the stock of forest products harvested from the land area

currently occupied by the timber stand. For our purposes we will assume that a

proportion of the harvested wood volume becomes a relatively long-lived product

(say lumber in housing) and that the remainder becomes a very short-lived product

(say toilet paper) and waste. We assume that the long-lived product decays at a

rate, θ, and that the carbon in the short-lived product and waste is released to the

atmosphere immediately upon harvest. We make this assumption so that we can

deal with a single pool representing products. We will use Z[t] to represent the

stock of carbon (tC/ha) in the long-lived product pool at time t: Z[0] will represent

11

the initial stocks.

Z[t] = e−tθZ[0] (16)

The amount of carbon emitted as a result of decay from the product pool is

Z′[t] = −θZ[0]e−θt (17)

At the time of harvest, a portion of the merchantable volume, γλV[t] is transferred

to the product pool. Our landowner is paid for additions to product pool stocks

and pays for reductions. The constant λ represents the proportion of merchantable

volume that enters the product pool: γ is a conversion factor used to convert mer-

chantable volume (m3 to mass of carbon (tC)).

The NPV of the additions to and removals from the product pool is calculated

as

NPVz = γλPcV(T )e−ρT +

∫ T

0PCe−ρt ∂Z(t)

∂tdt (18)

= e−ρT PcγλV[T ] −

(1 − e−T (θ+ρ)

)PCθZ[0]

θ + ρ(19)

The RVGR for timber plus biomass, DOM, and product pools is shown below,

with the numerator and denominator for the timber plus biomass and DOM pools

substituted in.

ρ =Xtbd + Pc

(e−tθZ[0] + γλV ′[T ]

)Y tbd + PcγλV[T ]

(20)

Including the product pool has the effect of making the denominator larger, there-

fore leading to a tendency to shorten the rotation. However the numerator is also

larger (and dependent on the initial stocks of the product pool) which would tend

to lengthen the rotation, relative to the timber plus biomass and DOM model. The

12

overall direction of the change in optimal harvest age relative to the timber plus

biomass and DOM pools is indeterminate.

3. Numerical example

We present a numerical example to illustrate further the points made in the

comparative statics analysis. Most of the parameter values for this example are

from Asante et al. (2011) representing a lodgepole pine [(Pinus contorta Dougl.

var. latifolia Engelm. (Pinaceae)) stand in northeastern British Columbia, Canada.

All costs and revenues are expressed in Canadian dollars (CAD). We refer the

interested reader to Asante et al. (2011) for detailed description of the parameter

derivation: the timber yield, price, and cost information is appropriate for a lodge-

pole pine stand managed for lumber production in northeastern British Columbia

at the time this work was done.

Merchantable timber volume (m3/ha) grows according to a Chapman-Richards

growth function

V[t] = 500.4(1 − e−0.027t)4.003. (21)

Biomass (tC/ha) also grows according in a Chapman-Richards function

B[t] = 198.6(1 − e−0.0253t)2.64. (22)

We use γ = 0.2 to convert between merchantable wood volume and carbon mass.

This is consistent with a carbon content of wood of approximately 200 kg m−3

(Jessome, 1977). We use the parameter, λ = 0.67, to express the proportion of

merchantable stand volume that is converted to lumber. This was calculated on

the basis of a lumber recovery factor of 250 board feet of lumber per cubic metre

of roundwood input.

13

The price of timber, Pw is set at 89.40 CAD/m3; volume based harvest cost, Cv,

is set to 47.55 CAD/ha; and area based harvest cost, Ca, is set to 6 250 CAD/ha.

Instead of setting establishment costs, Ce, to 1 250 CAD/ha, as in Asante et al.

(2011), we use Ce = 0 to avoid tangential discussion about the negative NPV that

would arise in the timber only case. For the purposes of the example, we use

carbon offset prices ranging from 1 to 50 CAD/tCO2e, which equates to 3.67 to

183 CAD/tC, using 3.67 as the ratio of the molecular weight of carbon dioxide to

the atomic weight of carbon.

We use a continuous discount rate, ρ = 0.05, a decay rate α = 0.00841, a

litterfall rate β = 0.01357, and a product decay rate θ = 0.00578. The derivation

of the numerical values for α and β are discussed in detail in Asante et al. (2011).

These parameters were estimated using a non-linear least squares estimation pro-

cedure to find the decay and litterfall rates that leads to the best approximation of

the projection of sum the non-living carbon pools generated by the CBM-CFS3

model developed by the Canadian Forest Service (Kurz et al., 2009), for a lodge-

pole pine stand in northeastern British Columbia following the timber yield curve

shown above. The product decay rate, θ, was estimated using non-linear regres-

sion to find the estimate of θ which resulted in the best fit to data tabulated by

Kurz et al. (1992) showing the proportion of orginal carbon remaining in lumber

over a 100 year time horizon in 20 year steps.

3.1. Timber only

The optimal harvest age considering only timber production for a single rota-

tion is 68.3 years and the associated NPV is 140 CAD/ha using the decision rule

in Eq. 3.

14

3.2. Timber plus biomass pool

Table 1 presents the optimal harvest age and associated NPV for a range of

carbon offset prices considering timber and the biomass carbon pool calculated

using the decision rule shown in Eq. 7. The optimal harvest age increases with

increasing carbon offset price: at 50 CAD/tCO2e the optimal decision is to never

harvest. NPV increases with increasing carbon offset prices from 140 CAD/ha

with a carbon price of 0 CAD/tCO2e to 4390 CAD/ha with a carbon price of 50

CAD/tCO2e. Note that all NPVs in Table 1 are greater than in the timber only

case.

3.3. Timber plus biomass and DOM pools

Tables 2 and 3 present the optimal harvest ages and associated NPVs for com-

binations of carbon price and initial DOM stocks. The optimal harvest age in-

creases with increasing carbon price, and decreases with increasing initial DOM

stocks. It is interesting to examine the joint effects of carbon price and initial

DOM stocks on NPV (Table 3). When initial DOM stocks are low, NPV increases

with carbon price; when DOM stocks are high, NPV decreases with increasing

carbon prices. This is due to the payment required for declining DOM carbon

stocks due to decomposition (Fig. 1).

All the NPVs presented in Table 3 for initial DOM stocks of 0 or 100 tC ha−1

are greater than in the timber only case. For initial DOM stocks of 200 tC ha−1

or greater, the NPVs are lower. For initial DOM stocks of 300 tC ha−1 or greater

NPVs decline with increasing carbon price. For initial DOM stocks of 200 tC

ha−1, the behaviour is more complicated, as the NPV declines with increasing

carbon price up to 20 CAD/tCO2e, but shows an increase at 50 CAD/tCO2e.

15

This behaviour will be explained using Fig. 3 Between prices of 24 and 30

CAD/tCO2e, the NPV function switches from having a maximum at a finite rota-

tion age to having a maximum at an infinite rotation age.

3.4. Timber plus biomass, DOM, and product pools

Tables 4 and 5 present the optimal harvest ages and associated NPVs for com-

binations of carbon price and initial DOM and product pool stocks. In comparison

to the model where timber, biomass, and DOM are considered, optimal harvest

ages are younger: substantially so for carbon prices greater than 20 CAD/tCO2e.

The optimal harvest ages for carbon prices of 50 CAD/tCO2e are older, but finite,

in contrast. Optimal harvest ages and NPVs decline with increasing initial stocks

of the product pool.

4. Conclusions

We developed an analytical model of the economically optimal harvest age

of a timber stand considering timber value, and the value of CO2 capture and

emissions considering biomass, dead organic matter, and wood product pools.

The model presented here considers just a single rotation. We do this because, as

we show in the analysis, optimal harvest age depends on initial stocks of carbon

in the DOM and product pools, and these will change from one harvest cycle to

another. We also apply the model in a numerical example using parameters from

Asante et al. (2011).

The results from our comparative statics analysis show that

1. as compared to the timber only model, the timber plus biomass model re-

sults in older optimal harvest ages,

16

2. when changes in the stock of carbon in DOM are valued, a higher initial

stock of DOM will generally result in a younger optimal harvest age,

3. the effect on optimal harvest age of incorporating biomass and DOM pools

relative to the timber only solution or the timber plus biomass solution is

indeterminate, and

4. the effect on optimal harvest age of incorporating biomass, DOM, and prod-

uct pools relative to the timber only solution or the timber plus biomass and

DOM solution is indeterminate.

There are some interesting observations to be made about our numerical ex-

ample. In this example,

1. participation in the market when initial DOM stocks are 200 tC ha−1 or

higher would be unattractive to a landowner as the NPV is lower than the

timber only case. Combinations of high carbon prices and high initial DOM

stocks can result in negative NPVs,

2. in our example, including biomass and DOM stocks always resulted in an

older rotation age than the timber only model, and a younger rotation age

than the timber plus biomass model, and

3. in our example, incorporating the product pool leads to a reduction in the

optimal harvest age.

In our minds, the most important conclusion to draw from this study is that

the optimal harvest age for a stand of timber will depend on which carbon pools

are being accounted for. Considering only the living biomass in a forest stand will

generally lead to an older optimal rotation than if changes in the carbon stored in

snags, the forest floor, and soil (what we referred to dead organic matter) is taken

into account.

17

One result from this study that surprised us was that NPV always decreased

with increasing initial stocks of carbon in the DOM pool or the product pool.

In retrospect, this should not have been a surprise Because of decay, the non-

living carbon pools represent a source of greenhouse gases. In our model, the

proportional decay rate is constant: increased stocks of carbon in the DOM and

product pools would result in greater absolute emissions of greenhouse gases, all

other things being equal. In a carbon market where a decision maker pays for

reductions in the size of the carbon pools under control, stocks of carbon become

a liability. Pardoxically, there is a benefit associated with increasing carbon stocks,

and a cost associated with holding the carbon stocks. This cost is what drives the

shorter rotations found when carbon stocks in the non-living pools are considered.

It is important to keep this paradox in mind when developing policies or markets

related to the sequestration of carbon in forests.

18

References

Aber, J. D., Botkin, D., Melillo, J. M., 1978. Predicting effects of different har-

vesting regimes on forest floor dynamics in northern hardwoods. Can. J. Forest

Res. 8 (3), 306–315.

Armstrong, G. W., Phillips W. E. 1989. The optimal timing of land use changes

from forestry to agriculture. Can. J. Agr. Econ. 37, 125–134.

Asante, P., Armstrong, G. W., Adamowicz, W. L., 2011. Carbon sequestration

and the optimal forest harvest decision: A dynamic programming approach

considering biomass and dead organic matter. J. Forest Econ. 17, 3–17.

California Climate Action Registry, September 2009. Forest Sector Protocol.

Version 3.0. Accessed 2009-10-20.

URL http://www.climateactionreserve.org/wp-content/uploads/2009/03

Chang, S. J. 1998. A generalized Faustmann model for the determination of opti-

mal harvest age. Can. J. Forest. Res. 28, 652-659.

Englin, J., Callaway, J., 1995. Environmental impacts of sequestering carbon

through forestation. Climate Change 31(1), 67–78.

Enzinger, S., Jeffs, C., 2000. Economics of forests as carbon sinks: an Australian

perspective. J. Forest Econ. 6, 227–249.

Harmon, M., Marks, B., 2002. Effects of silvicultural practices on carbon stores in

Douglas-fir western hemlock forests in the Pacific Northwest, U.S.A.: results

from a simulation model. Can. J. For. Res. 32, 863–877.

19

Hartman, R., 1976. The harvesting decision when a standing forest has value.

Econ. Inq. 14, 52—58.

Hoen, H., 1994. The Faustmann rotation in the presence of a positive CO2 price.

In: Helles, F., Lindhl, J. (Eds.), Proceedings of the Biennial Meeting of the

Scandinavian Society of Forest Economics, Gilleleije, Denmark, November

1993: Scand. J. For. Econ.35, 278-287.

Hoen, H., Solberg, B., 1997. CO2-taxing, timber rotations, and market implica-

tions. In: Sedjo, R. A., Sampson, R.N., Wisniewski, J. (Eds.), Economics of

Carbon Sequestration in Forestry.

IPCC, 2000. Land use, Land-Use Change, and Forestry. Cam-

bridge University Press, Cambridge UK, available online at

http://www.ipcc.ch/ipccreports/sres/land use/.

Jessome, A. P., 1977. Strength and related properties of woods grown in canada.

Forestry Tech. Report no. 21, Eastern Forest Products Laboratory, Ottawa,

Canada.

Kaipainen, T., Liski, J., Pussinen, A., Karjalainen, T., 2004. Managing carbon

sinks by changing rotation length in European forests. Environ. Sci. Policy

7 (3), 205–219.

Krankina, O., Harmon, M., 2006. Forest Management Strategies for Carbon Stor-

age. Forests, Carbon and Climate Change: A Synthesis of Science Findings.

Oregon Forest Resource Institute, Portland, OR., Ch. 5, pp. 78–91.

Kurz, W. A., Apps, M. J., Webb, T. M., MacNamee, P. J., 1992. The carbon budget

20

of the Canadian forest sector: Phase 1. Inf. Rep. NOR-X-326, Forestry Canada.

Northern Forestry Centre, Edmonton, Canada.

Kurz, W. A., Dymond, C. C., White, T. M., Stinson, G., Shaw, C. H., Rampley,

G. J., Smyth, C., Simpson, B. N., Neilson, E. T., Tyofymow, J. A., Metsaranta,

J., Apps, M. J., 2009. CBM-CFS3: A model of carbon-dynamics in forestry and

land-use change implementing IPCC standards. Ecol. Model. 220(4), 480–504.

Plantinga, A., Birdsey, R., 1994. Optimal forest stand management when benefits

are derived from carbon. Nat. Res. Model. 8(4), 373–387.

Stainback, G. A., Alavalapati, J., 2002. Economic analysis of slash pine forest

carbon sequestration in the southern U.S. J. For. Econ. 8(2), 105–117.

van Kooten, G. C., Binkley, C. S., Delcourt, G., 1995. Effect of carbon taxes and

subsidies on optimal forest rotation age and supply of carbon services. Am. J.

Agr. Econ. 77 (2), 365–374.

21

Table 1: Optimal harvest age and net present value by CO2 offset price when the biomass pool isconsidered.

Carbon price Optimal harvest NPV(CAD/tCO2e) age (yr) (CAD/ha)

1 70.0 2112 71.8 2835 77.9 507

10 91.6 90220 235 1 75050 ∞ 4 390

Table 2: Optimal harvest age by CO2 offset price and initial dead organic matter stocks when thebiomass and dead organic matter pools are considered.

Carbon price Initial DOM stocks (tC ha−1)CAD/tCO2e 0 100 200 300 400

1 69.6 69.5 69.3 69.2 69.12 70.9 70.6 70.4 70.1 69.85 75.1 74.4 73.7 73 72.3

10 83.3 81.8 80.3 78.8 77.220 108 104 99.9 95.9 91.950 ∞ ∞ ∞ ∞ ∞

1

Table 3: Optimal net present value by CO2 offset price and initial dead organic matter stocks whenthe biomass and DOM pools are considered.

Carbon price Initial DOM stocks (tC ha−1)CAD/tCO2e 0 100 200 300 400

1 236 184 132 80.6 28.72 333 229 125 21.3 −82.55 627 366 106 −154 −415

10 1 130 605 82.0 −441 −96320 2 170 1 120 63.4 −989 −2 04050 5 400 2 770 125 −2 510 −5 150

2

Table 4: Optimal harvest age by CO2 offset price, initial dead organic matter stocks, and initialproduct pool stocks when the biomass, DOM, and product pools are considered.

Carbon price Initial DOM Initial product pool stocks (tC ha−1)CAD/tCO2e stocks (tC ha−1) 0 50 100 150 200

1 0 69.5 69.4 69.4 69.3 69.2100 69.3 69.3 69.2 69.2 69.1200 69.2 69.1 69.1 69.0 69.0300 69.1 69.0 69.0 68.9 68.8400 68.9 68.9 68.8 68.8 68.7

2 0 70.6 70.5 70.4 70.3 70.2100 70.4 70.2 70.1 70.0 69.9200 70.1 70.0 69.9 69.8 69.6300 69.8 69.7 69.6 69.5 69.4400 69.6 69.4 69.3 69.2 69.1

5 0 74.2 74.0 73.7 73.4 73.1100 73.6 73.3 73.0 72.7 72.4200 72.9 72.6 72.3 72.0 71.7300 72.2 71.9 71.6 71.3 71.1400 71.5 71.2 70.9 70.7 70.4

10 0 80.9 80.3 79.7 79.1 78.5100 79.5 78.9 78.3 77.7 77.1200 78.1 77.5 76.9 76.3 75.7300 76.7 76.0 75.4 74.8 74.2400 75.2 74.6 74.0 73.3 72.7

20 0 97.8 96.4 94.9 93.5 92.1100 94.6 93.1 91.7 90.3 88.9200 91.3 89.9 88.4 87 85.6300 88.0 86.6 85.1 83.7 82.2400 84.6 83.2 81.7 80.3 78.8

50 0 281 268 257 241 227100 261 247 232 218 204200 239 220 209 194 180300 214 198 183 169 156400 186 171 157 144 133

3

Tabl

e5:

Opt

imal

netp

rese

ntva

lue

byC

O2

offse

tpri

ce,i

nitia

ldea

dor

gani

cm

atte

rsto

cks,

and

initi

alpr

oduc

tpoo

lsto

cks

whe

nth

ebi

omas

spo

olis

cons

ider

edZ

0PC

O2

DO

M0

050

100

150

200

10

238

219

201

182

164

100

186

167

149

130

112

200

134

116

97.0

78.4

59.8

300

82.3

63.7

45.2

26.6

7.96

400

30.5

11.9

−6.

71−

25.3

−43.9

20

336

299

262

225

187

100

232

195

158

121

83.5

200

129

91.4

54.1

16.9

−20.3

300

24.8

−12.5

−49.7

−86.9

−12

440

0−

79.0

−11

6−

153

−19

1−

228

50

634

541

448

354

261

100

374

281

187

93.8

0.46

020

011

420.3

−73.0

−16

6−

251

030

0−

147

−2

310

−33

3−

426

−5

110

400

−40

7−

491

0−

593

−68

6−

779

100

114

095

376

557

738

910

061

74

210

242

54.5

−13

320

094.7

−92.8

−28

0−

468

−65

530

0−

428

−61

5−

802

−98

9−

118

040

0−

949

−1

140

−1

320

−1

510

−16

100

200

218

01

800

142

01

050

668

100

113

075

037

3−

4.90

−38

220

077.1

−30

0−

678

−1

050

−1

430

300

−97

3−

135

0−

173

0−

210

0−

248

040

0−

202

0−

2310

0−

277

0−

315

0−

353

050

05

400

446

03

510

256

01

610

100

277

01

820

865

−84.8

−1

030

200

125

−82

5−

177

0−

272

0−

367

030

0−

251

0−

346

0−

441

0−

536

0−

631

040

0−

515

0−

610

0−

705

0−

800

0−

895

0

4

Figure 1: Optimal harvest decision rule and trajectory of DOM. Reprinted from Asante et al. 2011.

5

60 80 100 120 140Age HyearsL

0.02

0.04

0.06

0.08

0.10Rate

Figure 2: First order condition for optimization of harvest age. The horizontal line representsa discount rate of 5%. The curve is the relative value growth rate. The intersection of the tworepresents the optimal harvest age, 68.3 years, in this example.

6

20

20

40

40

60

60

80

80

100

100120

120

140

0 10 20 30 40 50

60

80

100

120

140

160

180

200

Price H$�tCO2eL

Har

vest

age

HyrL

Figure 3: Contour plot of net present value function for timber plus biomass and DOM pools forinitial DOM stocks of 200 tC ha−1.

7